Challenges in Fluid Dynamics: A New Approach

245 Pages·2017·6.227 MB·English

by  R.Kh. Zeytounian (auth.)

R.Kh. Zeytounian
Challenges
in Fluid
Dynamics
A New Approach
Challenges in Fluid Dynamics
R.Kh. Zeytounian
Challenges in Fluid
Dynamics
A New Approach
R.Kh.Zeytounian
Universite´deLilleI
LML–cite´Scientifique
Villeneuved’Ascq,France
ISBN978-3-319-31618-5 ISBN978-3-319-31619-2 (eBook)
DOI10.1007/978-3-319-31619-2
LibraryofCongressControlNumber:2017962965
©SpringerInternationalPublishingAG2017
Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpartof
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or
dissimilarmethodologynowknownorhereafterdeveloped.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publicationdoesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexempt
fromtherelevantprotectivelawsandregulationsandthereforefreeforgeneraluse.
Thepublisher,theauthorsandtheeditorsaresafetoassumethattheadviceandinformationinthis
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained
hereinor for anyerrors oromissionsthat may havebeenmade. Thepublisher remainsneutralwith
regardtojurisdictionalclaimsinpublishedmapsandinstitutionalaffiliations.
Printedonacid-freepaper
ThisSpringerimprintispublishedbySpringerNature
TheregisteredcompanyisSpringerInternationalPublishingAG
Theregisteredcompanyaddressis:Gewerbestrasse11,6330Cham,Switzerland
Thisbookisdedicatedtothememoryof:Arto
MkhitarianS.N.MergelyanandI.A.Kibeland
also:myfriendandcolleagueJean-Pierre
Guiraud
Preface
Zeytounian’s(2012)bookmakesagoodcasethathisRAMapproachdeservesmoreserious
computational attention than it has received. His presentation is generally clear, though
SpringershouldhavedevotedmoreefforttocleaninguptheEnglishandthetypography.
RobertE.O’Malley,Jr—BooksReviewsofSIAMReview54(3)page613
ThemostoutstandingChap.4ofZeytounian’s(2014)bookdevelopsthebasicNavier-
Stokes–Fourier (NSF) equations in a historical context. Overall, readers will realize the
unique value of Zeytounian’s work and perspective and will come to appreciate his
willingnesstotackleverydifficultproblems,aimingtohelpFluidDynamicalNumericians.
Anexampleismodelingturbo-machineryusingthenumberofbladesonarotorasalarge
parameter.
RobertE.O’Malley,Jr—BooksReviewsofSIAMReview56(3)page569
MylasttwobookswerepublishedbySpringer-VerlaginHeidelbergin2012and
2014. As a research monograph, the 2012 book is mainly devoted to the Navier-
Stokes-Fourier (NSF) system of equations for a Newtonian fluid in the case of
compressible, viscous, and heat-conducting flows. The main tool developed there
was what I call the Rational Asymptotic Modelling (RAM) approach. In this 2012
book,themainobjectivehasthreefacets.First,to“deconstruct”theNSFsystemof
equations in order to bring some unity to the various partial, simplified, and
approximatemodelsusedinclassicalNewtonianfluiddynamics;thisfirstfacetof
themainobjectiveisobviouslyquiteachallengeandhasveryimportantpedagog-
icalconsequencesforuniversityeducation.Thesecondfacetofthemainobjective
istooutlineaconsistentrationalasymptotictheoryformodelingfluidflowsonthe
basis of a typical NSF initial and boundary value problem. The third facet is an
illustrationofourrationalasymptoticmodeling(RAM)approachforvarioustough
technological and geophysical problems from aerodynamics, thermal and
thermocapillaryconvections,andalsometeo-fluiddynamics.
Concerningthe2014book—a“ScientificAutobiography”—thereaderisguided
through my somewhat unconventional career: first discovering fluid mechanics,
and then devoting more than fifty years to intense work in the field. Using both
personalandgeneralhistoricalcontexts,thisaccountwouldbeofbenefittoanyone
vii
viii Preface
interested in the early and contemporary development of an important branch of
theoreticalandcomputationalfluidmechanics.
The present book (written during the years 2013–2016), together with the two
mentioned above, is a major part of a “trilogy” providing a challenging modern
point of view of fluid dynamics that examines the origins of working models and
theirclassification.Thislecturecourseseemstocorrespondwelltoasuggestionof
PaulGermain(seetheconclusioninhislastpaper[1])concerning“fluidmechanics
inspiredbyasymptotics”asanewwaytolookatproblems,whichhasprovenitself
veryusefulingasdynamics,hyposonic,supersonic,andhypersonicaerodynamics,
magnetofluid dynamics, progressive waves, internal structure of shock waves,
hydrodynamic instability, combustion with high activation energy, waves near
breaking,atmosphericmotions,convectioninliquids, nonlinearacoustics,andso
on! More precisely then, in the present book Challenges in Fluid Dynamics—a
close companion to the 2012 book—the reader will find many new facets and
applicationsofourrationalasymptoticmodelling(RAM)approach.
Inparticular,mypurposehasalsobeentoprovidealogicallyconsistentanswer
to two questions mentioned below concerning the curious “fragmentation” of
Newtonianfluiddynamics,whichgeneratesanever-ending andvaried listoffluid
dynamicsproblemsdescribingcertainveryparticularmodelsoffluidflows!These
twoveryreasonablequestionswereraisedinabookwrittenover40yearsagoby
MartinShinbrot[4]:
Whereishetobegin[concerning,thenever-endinglistofparticularfluidflowmodels],and
whichoftheseisfundamental,andwhichcreatedtoaidintheexplicationofsomespecific
physicalproblem?
In reality, it seems to me that today mostfluid dynamicists and applied mathe-
maticiansonlylearntheanswerstothesetwoquestionsaftertheyhaveworkedona
specificfluiddynamicsproblempresentedtothemonanadhocbasis.Thefirststep
hereisthedimensionlessanalysisofNSFproblemswithspecificnon-dimensional
parameters relating to the various physical effects and associated with the
corresponding terms in NSF problems. The second step involves the emergence
fromthesedimensionlessNSFproblemsofalargefamilyofapproximaterational
reduced working models, which leads to a classification according to limiting
values (mainly zero or infinity) of a large number of dimensionless reduced
parameters(inNSFproblems):Reynoldsnumber(viscosity),Machnumber(com-
pressibility), Prandtl number (heat conduction), Strouhal number (unsteadiness),
Froude/Boussinesq numbers (gravity), Rossby/Kibel numbers (earth rotation),
Grashof/Rayleigh numbers (thermal convection in liquids), Marangoni number
(thermocapillaryeffect),andsoon.
In this way, we also get a clear view of the origins of the emerging reduced
working models, and this gives another understanding of the NSF equations, and
undoubtedlyanewstatustotheStokes“fluidity”concept,discussedinSect.1.3,as
abasisforclassicalNewtonianfluidmechanics.Ithink,ontheonehand,thatthese
emerging approximate rational reduced working models, derived using the RAM
approach, should also be very welcome for “pure” mathematicians, when they
Preface ix
discover that there are rigorous results concerning the existence, regularity, and
uniquenessofsolutionsofvariousinitial-boundaryvalueproblemsinfluiddynam-
ics! This seems to me obvious if we take into account the fact that, in so-called
“mathematical fluid dynamics”, theconsidered fluidflowproblemsare very often
in no way rigorously linked with the fundamental classical NSF equations for
Newtonianfluidflows!
At the same time, the approximate rational and consistent reduced working
models we derive from very stiff NSF problems should also be valuable for
numericiansdealingwithnumericalsimulations/computationsusinghigh-memory
high-speedcomputers.
In a certain sense, Chap. 4 of our 2014 book, devoted to the interrelationship
between NSF equations and our RAM approach, can be considered as a short
preliminary presentation of the NSF system of equations via the RAM approach,
in the space of just 54 pages. And in Sect 3.4, the reader will find a short, but
neverthelessilluminatingaccountoftheRAMapproach,througharatherlaborious
applicationtothederivationandjustificationofsomeapproximatereducedworking
meteo-atmosphericmodelsfortheweatherforecast.
Obviously,aZeytounian“trilogy”isbasicallyverydifferentfromatraditional
courseoflecturesonfluiddynamics,andgivesadifferent,ratherchallengingview
offluiddynamics,inwhichtheNSFsystemofequationsisgiventhemostimportant
place!Indeed,themainnoveltyinourapproachcomeschieflyfromthefactthatthe
fundamental system of NSF equations governing the Newtonian motion of a com-
pressible, viscous, and heat conducting fluid flow is given a central role in our
discussion. This leads to a new vision of the universality of NSF equations as a
fundamentalmathematicalsystemforthestudyofmanyaspectsofNewtonianfluid
dynamics,butalsotoits“deconstruction”viatheRAMapproach!
However,astrongrestrictionisassumed(atleastatthepresenttime!),namely,
thattheconsideredfluidflowsarelaminar!Thisisunfortunatebecausesomefluid
flows appear to be very tumultuous, changing rapidly from point to point or from
time to time at a given position in 3D time-space! Such non-laminar and very
tumultuous fluid flows, so-called turbulent fluid flows (where the usual laminar
motion becomes turbulent if for instance the Reynolds numbers is too high and
there is a laminar portion upstream, followed by a transition to turbulence) defi-
nitelyfalloutsidethescopeofourconsiderationsinthepresentbook.Discussionof
turbulentmotionisnecessarilyaverylongbusiness,butthereadercanfindvarious
valuable books dealing with turbulence, despite the fact that turbulence is a huge
subject of intensive ongoing research. It seems to me that, in the book by Sagaut
etal.[3],thereadercanfindacomprehensivedescriptionofmodernstrategiesfor
turbulentflowsimulation,rangingfromturbulencemodellingtothemostadvanced
numericalmethods.
Asafinalremarkweobservethataconsequenceofourmacroscopic-continuum
approach, equations of state and transport coefficients (such as viscosity and heat
conduction)aregivenasphenomenologicalorexperimentaldataandarenotrelated
tomicroscopicequations,i.e.,lawsgoverningmolecularinteractions.
x Preface
InGolse[2],apaperwith142pages,thecuriousandmotivatedreaderwillfinda
good account of the Boltzmann equation and its hydrodynamic limit However,
Golse mentions (on p. 238 of [2]) rather disappointedly that: “The derivation of
theNSFsystemofequationsfromtheBoltzmannequationleadstodissipationterms
thatareoftheorderoftheKnudsennumber(Kn(cid:1)1;ratioofMach/Reynolds),and
thereforevanishinthehydrodynamiclimit,Kn!0.Inotherwords,theNSFsystem
ofequationsisanasymptoticexpansionoftheBoltzmannequationintheKnudsen
number,andnotalimitthereof!”
IstressagainthatthedeconstructioncarriedoutintheRAMapproachisbased
onarationallyargued,consistent,non-ad-hoc,andnon-contradictoryapproachto
the full unsteady NSF system of equations for modelling Newtonian fluid flows,
tackled in the spirit of asymptotics. Our main goal is the modelling, and not the
search for solutions to the derived reduced working model equations with initial
and boundary conditions. The aim is thus to assist numericians in the context of
theirsimulation/computationsusingsuper-high-powercomputers!
ItseemstomethattheRAMapproachopensupnewvistasforthederivationof
manyvaluableworkingmodelproblems.Suchmodelsofferalargepaneloftools
forfluiddynamicistsinterestedinengineering,environmental,meteorologicaland
atmospheric,orgeophysicalapplications.Itgivestheoreticallyorientedresearchers
a way to exchange with numericians expert in high-speed computing, in order to
provide them with models that are much less stiff than “brute force”, starting
directlyfromthefulldimensionlessunsteadyNSFsystemofequationsandbound-
aryconditions!
Letusendthisprefacewithabriefdescriptionoftheninechaptersinthepresent
book,followingtheintroduction.Chapter1describessomehistoricalstepsrelating
to the discovery of the NSF equations, i.e., from Newton (1686) to Stokes(1845),
viaEuler(1755),Navier(1821),Cauchy(1828),Fourier(1833),andSaint-Venant
(1843).Ihopethatthischapterwillstimulatetheinterestofmanyreaders,andwhy
not also motivate some young mathematicians, to carry out theoretical investiga-
tionsofmorerealisticappliedproblemsinvolvingNewtonianfluidflowsgoverned
bytheNSFequations!
In Chap. 2, the reader can find first a formulation of a typical initial-boundary
value, unsteady NSF problem in the case of a thermally perfect gas, and then the
NSF equations applicable to an expansible liquid, Navier-Stokes (NS, physically
unreal,butoftenusedbymathematicians)barotropiccompressiblefluidflow,and
NSFequationsappliedinnonlinearacousticsasabranchoffluiddynamics.
InChap.3,webrieflypresentthefoundationsoftheRAMapproachtoempha-
sizeourmainpostulateandsomekeystepsinitsrealization,relatingprincipallyto
thedimensionlessapproach,asymptotics,andvariouslimitingprocessesrelativeto
some of the most significant reduced dimensionless parameters in the NSF equa-
tions and boundary conditions. The case of dimensionless NSF equations for
atmosphericmotions(Sect.3.4)iscarefullyconsideredtoillustratethepossibilities
oftheRAMapproach.
Preface xi
InChap.5,Ihavetentativelysketcheda“theoryofmodels”emergingfromthe
full unsteady dimensionless NSF problem in the RAM approach. In Figs. 5.1 and
5.2,thereaderwillfindmanyvaluableapproximateworkingmodels,whichgivea
preliminaryideaabouttheoriginsandclassificationofafamilyofmodelsarising
asaRAMlimitofthedimensionlessNSFproblem.Inparticular,Fig.5.2showsthe
approximate reduced workingmodelsforhighandlowReynoldsnumbers(Re(cid:3)
1andRe(cid:1)1),andalsoforlowMachnumber(M(cid:1)1).
Chapter 6, is devoted to the RAM “deconstruction/modelling” of four typical,
but very stiff NSF problems, namely, Re (cid:3)1 and M (cid:1) 1, meteo-fluid dynamics
models, and models for a weakly expansible liquid heated from below. The
approximate working models derived for the above-mentioned typical cases are
sketchedinFigs.6.1to6.4.
Chapter 8 is a miscellany of major applications of the RAM approach I have
beeninvolvedinoverthepast50years(1960–2010)(seeSect.6.1).Chapters4,7
and9aredevotedtosomeconcludingremarksaboutChaps.2,3,5,6,and8.
I have written the present book as the last, coming at the end of 40 years of
writing, from 1974 to 2014, which has resulted in 14 books, mainly devoted to
rational modelling of Newtonian fluid flows. I hope that the above-mentioned
“trilogy” will be useful to students in the last years of university and young
researchers working in the field of fluid dynamics and motivated by modelling,
numericalanalysis,andsimulation/computationofverystiffcomplexrealtechno-
logicalandgeophysicalflows.Our“trilogy”shouldalsoproveusefulasareference
forscientistsandengineersworkinginthevariousfieldsoffluiddynamics.
IowethanksfirsttoDr.ChristianCaron,executivepublishingeditorforphysics
at Springer, for his continued support during many years in the edition of my
various books, and who suggested to Dr. Jan-Philip Schmidt, Springer editor for
interdisciplinaryandappliedsciences,thepublicationofthepresent“Challenges”
intheappliedmathematicsandengineeringprogram.Asformypreviousbooks,I
had several fruitful discussions with J. P. Guiraud, whose help was decisive in
renderingmyownideasmoreprecise!
My sincere appreciation goes to Dr. Jan-Philip Schmidt, Springer DE, and his
copyeditingandproductionteamfortheirprofessionalandkindassistance.
Paris/Yport R.Kh.Zeytounian
October2015
References Cited in the Preface
[1]GermainP(2000)The“New”mechanicsoffluidsofLudwigPrantdl.In:MeierG(ed)Ludwig
PrantdleinFührerinderStromungslehre.Vieweg,Braunschweig,pp31–40
[2] Golse F (2005) The Boltzmann equation and its hydrodynamic limits. In: Dafermos CM,
Feireisl E (eds) Handbook of differential equations, evolutionary equations, vol. 2, chap.
3.ElsevierB.V,pp159–301
[3]SagautP,DeckS,TerracolM(2006)Multiscaleandmultiresolutionapproachesinturbulence.
ImperialCollegePress